Better Search Took This AI Agent From 36% to 98% Accuracy - Aamir Shakir, Mixedbread

Makes the case that the OCR, chunk and embed pipeline loses information that agents need. Multimodal late-interaction retrieval is presented as the fix, with a reported accuracy gain from 36% to 98%.

• chunking — Splitting documents into passages small enough to embed and retrieve individually — the step that quietly determines whether retrieval works at all.
• multimodal — A model that works with more than text — reading images, audio, or video, and sometimes generating them too.

• videoYour AI Agents Shouldn't Care Where They Run - Nikhil Harithas, Factory
• videoMake Your AI Agents Predictable With Graph Engineering - Zachary Blumenfeld, Neo4j
• videoYou Can't Tell What's Real Anymore. Here's What Comes Next - Cristóbal Valenzuela, Runway
• videoYour Company Should Train Its Own AI Models - Will Brown, Prime Intellect

• videoVector Isn't Enough: Hybrid Search & Retrieval — Jeff Vestal & James Williams, ElasticAI Engineer
• videoGrep or Embeddings? Agentic Search Over Company Documents — George He, LlamaIndexAI Engineer
• videoAgentic Search vs Vector Search for Coding Agents: We Ran the Eval — BraintrustAI Engineer

Checking sign-in…

Loading comments…